Estrogen deficiency is known to result in deterioration of the skeletal and cardiovascular systems in postmenopausal women. Osteoporosis is characterized by a progressive decrease in bone density which can lead to an increased incidence of bone fractures. This condition results when the rate of bone resorption exceeds that of bone formation. Several disorders induce abnormalities in bone remodeling, the most common of which is loss of gonadal steroid action, as can occur in menopause or in male or female hypogonadism. Thus, bone loss is not itself a disease but rather is a consequence of endocrine imbalance.
Several types of therapeutic agents are available or in development for use when estrogen replacement therapy (ERT) is indicated for prevention of postmenopausal bone loss. The most important of these suppress bone resorption and formation in a manner that maintains net bone balance, presumably as a consequence of interaction with estrogen receptors (ER) in osteoblasts and osteoclasts, cell types responsible for bone maintenance. Thus, 17.beta.-estradiol and its orally active analogs, including conjugated equine estrogens, are widely used in ERT in postmenopausal women. The use of these steroids in ERT can, however, cause serious adverse effects on the reproductive system. For example, administration of 17.beta.-estradiol is associated with increased risk of developing uterine cancer and endometriosis; there is also a possible link between estrogen use and breast cancer.
Steroidal estrogen substitutes such as ethynyl estradiol and mestranol have been used in ERT, but are also disfavored. These compounds have been associated with a number of adverse side effects including myocardial infarction, thromboembolism, cerebrovascular disease, and endometrial carcinoma. Fortunately, however, the estrogen receptor has been found to bind not only estradiol and other steroidal compounds but also a diverse array of aromatic nonsteroidal structural types, exemplified by mono- and dihydroxylated triarylethylenes. This observation has stimulated a significant amount of research in an effort to identify effective nonsteroidal compounds for use in ERT. Agents displaying bone-selective estrogenicity are of particular interest due to their potential for reduced reproductive tract toxicity compared with conventional estrogens. M. Sato et al., FASEB J. 10: 905-912 (1996); H. Ke et al., Endocrinology (Baltimore) 136: 2435-2441 (1995).
Some nonsteroidal estrogen antagonists have, somewhat surprisingly, showed promise in preventing bone loss in postmenopausal women. An example of such a nonsteroidal antiestrogen is tamoxifen (TAM), ((Z)-2-[4-(1,2-diphenyl-1-butenyl)phenoxy]-N,N-dimethylethanamine), which is a triphenylethylene derivative. Tamoxifen effectively antagonizes the growth-promoting effect of estrogens in primary target tissues such as the uterus and ovary. Tamoxifen is currently marketed for the treatment of breast cancer, and is also administered to initiate ovulation in anovulatory women. Adverse side effects, however, can include reproductive tract effects such as endometriosis and endometrial cancer (M. Killackey et al., Cancer Treat. Rep. 69: 237-238 (1985); M. Seoud et al., Obstet. Gynecol. 82: 165-169 (1993)). Clomiphene (CLO) (2-[4-(2-chloro-1,2-diphenylethenyl)phenoxy]-N,N-diethylethanamine), which is structurally close to tamoxifen, is another nonsteroidal antiestrogenic pharmaceutical compound that has been used in ERT. The preparation of clomiphene is described in U.S. Pat. No. 2,914,563. Clomiphene is prescribed to induce ovulation in infertile women with physiological indications of normal estrogen levels. In the hypothalamus, clomiphene antagonizes estradiol-mediated feedback inhibition of gonadotrophin-releasing hormone secretion.
Interestingly, although they antagonize the growth-promoting effect of estrogens in primary reproductive target tissues, the nonsteroidal antiestrogens CLO and TAM prevent development of osteopenia in the ovariectomized (OVX) rat to a degree approaching that of 17.beta.-estradiol. Histomorphometric analysis of bone specimens from OVX rats receiving 17.beta.-estradiol, or the nonsteroidal antiestrogens tamoxifen (TAM) or clomiphene (CLO) has shown a decreased rate of bone turnover and maintenance of normal bone mass for all three experimental groups compared to that observed in untreated OVX rats (M. Jimenez et al. Endocrinology 138:1794-1800 (1997); L. Moon et al., Endocrinology 129:1568-1574 (1991); T. Wronski et al., Endocrinology 123:681-686 (1988). Likewise, estrogen or TAM administration to OVX rats resulted in decreased serum levels of osteocalcin (D. Williams et al., Bone Mineral. 14:205-220 (1991)). Osteocalcin is a bone matrix protein which is released into the serum during bone formation, thus serving as a specific indicator of bone turnover. TAM and CLO have also been shown to prevent bone loss in postmenopausal women (R. Love et al., Breast Cancer Res. Treat. 12:297-302 (1988); R. Young et al., Int. J. Fertil. 36:167-171 (1991)). In contrast to 17.beta.-estradiol, however, CLO and TAM are only moderately uterotrophic. U.S. Pat. No. 4,894,373 to Young describes the use of clomiphene, tamoxifen, nafoxidene, and other antiestrogenic compounds in the treatment of menopause and osteoporosis. However, despite results suggesting that these compounds are less estrogenic in reproductive tissues than steroidal estrogens, the adverse reproductive tract effects of tamoxifen, raloxifene, and other nonsteroidal antiestrogens (V. Jordan, Pharmacol. Rev. 36:245-276 (1984), T. Willson et al., Endocrinology 138:3901-3911 (1997)) are problematic.
Nonsteroidal estrogenic compounds are also of interest in the continuing effort to improve ERT. However, the use of nonsteroidal estrogenic compounds, like estradiol, in ERT is expected to be accompanied by a detrimental effect on the reproductive tract. For example, compounds that are estrogenic (growth-promoting) in MCF-7 cells, such as diethylstilbestrol (DES) and chlorotrianisene, are known to cause undesirable uterotrophic effects in the OVX rat (P. Ruenitz et al., J. Steroid Biochem. Mol. Biol., 63, 203-209 (1997); M. Shelby et al., Environ. Health Perspect., 104, 1296-1300 (1996)).
An estrogen mimetic that showed initial promise for use in ERT, based upon findings that suggested it could be selectively estrogenic in nonreproductive tissues (P. Ruenitz et al., J. Med. Chem. 39:4853-4859 (1996)) was 4-hydroxytamoxifen acid, a nonsteroidal metabolite of tamoxifen (4HTA; (E,Z)-2-{4-[1-(p-hydroxyphenyl)-2-phenyl]-1-butenyl}phenoxyacetic acid). 4HTA was shown in U.S. Pat. No. 5,189,212 to have estrogenic activity, a result that was quite unexpected since both the parent compound tamoxifen and the related compound clomiphene have an opposite, antiestrogenic effect in vivo. ER affinity, estrogenic (i.e., growth stimulatory) potency and estrogen efficacy were compared for a group of synthetic monophenolic triarylethylene acetic acids and analogs that included 4HTA (P. Ruenitz et al., J. Med. Chem. 39:4853-4859 (1996)). These synthetic compounds were designed to evaluate the importance of structural features known or anticipated to facilitate ER affinity. 4HTA was shown to have high ER affinity and strong growth stimulatory potency (estrogenicity) in MCF-7 breast cancer cells. In addition, 4HTA functioned as a partial agonist in stimulating growth (79% maximal growth-stimulatory effect, as a percent of that of estradiol) in the MCF-7 cell proliferation assay, while exhibiting weak antagonist potency in an inhibition assay in the presence of estrogen (P. Ruenitz et al., J. Med. Chem. 39:4853-4859 (1996); S. Wilson et al., J. Steroid Biochem. Molec. Biol. 42:61314 616 (1992)). It was also reported that 4HTA exhibited an effect on trabecular bone maintenance that was qualitatively similar to that of estradiol, yet that it had no observable uterotrophic effect. Taken together, these characteristics suggested the possibility of differential estrogenicity for 4HTA; however we subsequently discovered, as disclosed herein, that 4HTA is moderately uterotrophic and that it does not have a bone protective effect.
A saturated analog of 4HTA, 4-[1-(4-hydroxyphenyl)-2-phenylethyl]phenoxyacetic acid (HPPA), was also evaluated by P. Ruenitz et al. (J. Med. Chem. 39:4853-4859 (1996)). In contrast to its parent compound, HPPA exhibited unpromising ER affinity. However, despite its relatively low ER affinity, HPPA exhibited estrogenic potency approaching that of 4HTA in MCF-7 cells. HPPA also was a full agonist in stimulating growth (102% maximal growth-stimulatory effect, as a percent of that of estradiol) and did not function as an estrogen antagonist in the growth inhibition assay. Growth-promoting effects of both 4HTA and HPPA were fully antagonized by the antiestrogen tamoxifen, suggesting that such effects are mediated via ER (P. Ruenitz et al., J. Med. Chem. 39:4853-4859 (1996)). That report did not investigate the level of extra-reproductive tract estrogenicity, if any, of the compounds under study.
What is needed for use in ERT is an estrogen mimetic having selective estrogenicity which, like estrogen, counteracts the skeletal and cardiovascular deterioration that often accompanies menopause, but lacks the reproductive tract effects typically associated with currently available estrogens and antiestrogens.